How do we audit a remote-proctoring tool deployment for disability disparate impact?
You audit it by measuring flagging or scoring outcomes across relevant cohorts — including students using accommodations — on a fixed, representative probe set, and documenting any disparity and your mitigation. Reproducible, comparable cross-cohort…
Answer.
You audit it by measuring flagging or scoring outcomes across relevant cohorts — including students using accommodations — on a fixed, representative probe set, and documenting any disparity and your mitigation. Reproducible, comparable cross-cohort measurement is the core evidence. Planisphere produces exactly that: a sha-pinned record of model behaviour across cohorts, run without needing student PII where designed; it surfaces disparate impact for your review, it does not certify §504/ADA compliance.
The mark behind the answer.
automated proctoring/grading · §504 disparate-impact across cohorts.
More on Proctoring Fairness.
Try Proctoring Fairness with a free test key.
Start with the free sandbox. Planisphere measures model behaviour and emits a reproducible, sha-pinned record — it does not certify, file, or give legal advice.